Four the past four billion years or so the continents have ambled back and forth across the face of the Earth, sometimes forming up to form vast supercontinents such as Pangea or Rodinia, at other times going on their own ways. For much of the past five hundred million years these continents have been covered by vegetation.
Both of these have had profound effects upon the climate. The position of the continents effects the flow of the oceans, and therefore the atmosphere; for example the Pleistocene ice ages are believed to have started after a land bridge developed between North and South America, preventing the flow of water between the Atlantic and the Pacific, and diverting the Global Ocean Conveyer-belt through the waters surrounding Antarctica. Vegetation effects cloud cover by pumping water into the atmosphere via transpiration, the process by which plants suck up water from the ground via their roots then let it evaporate from their leaves, driving their circulatory systems.
This month Esther Sanromá and Enric Palle of the Instituto de Astrofísica de Canarias published a paper on the online arXiv database at Cornell University Library detailing the results of an attempt to accurately model the cloud cover of the Earth at various points in its history. The theory behind this is fairly simple; a model is created in which the surface of the Earth is divided up into a large number of cells, each of which has a weather patter, which can be influenced by conditions acting on the cell (heat received from the sun, the nature of the ground cover bellow etc.) as well as the weather in neighboring cells. The maps used were based upon those from Ron Blakey of Northern Arizona University's Department of Geology's Global Paleogeography Website.
Quite accurate models of the modern climate can be made in this fashion, but for ancient Earths the task was going to be harder. Accurate models require modeling vegetative ground cover to understand the effect this has on the climate; thus tropical rain forests, high latitude boreal forests and grasslands have to be treated differently. Unfortunately we don't have a good enough understanding of the vegetative cover over the geological timescale to make this possible. Instead Sanromá and Palle were forced to use a far simpler model of ground cover; either vegetated or desert with desert further divided by latitude.
This had clear limitations, but it was thought that if a reasonably accurate model of the modern Earth could be built up in this way, then it would be worth proceeding with the models of ancient climates. The modern Earth has a photometric variability of 3.29% and a mean albedo of 0.315 (that is to say on average 31.5% of the light that falls onto the Earth is reflected back into space, but that this average can vary by up to 3.29%). The computer simulation was able to create a model with cloud cover of 4.16% and an albedo of 0.325; not an exact match for our climate, but reasonably close.
Having thus calibrated the simulation Sanromá and Palle moved on the model the climate of the Late Cretaceous, 90 million years ago. At this time the ancient continent of Pangea had completely broken up, and the continents had yet to start to collide again, so that they remained as separate entities scattered around the globe. In addition the Late Cretaceous had a much warmer climate than today, so that it lacked ice caps. This lead to the creation of vast inland seas on most of the continents, further breaking up the Earth's land cover. Armed with this information Sanromá and Palle came up with a model which gave the Late Cretaceous Earth a photometric variability of 4.27% and an albedo of 0.331 - not greatly different from that of the modern Earth.
The Earth in the Late Cretaceous.
Next Sanromá and Palle built a model of the Earth in the Late Triassic, 230 million years ago. At this time the world's land masses were joined into a single supercontinent, Pangea, that reached almost from pole to pole. In the Late Triassic this supercontinent was starting to break up, the plates in the east had separated giving Pangea a 'C' shape, and forming a knew ocean, the Tethys. The Triassic had a hot, dry, climate with no glaciation at either pole. Sanromá and Palle's model of the Triassic gave a photometric variability of 5.02% and an albedo of 0.327; still comparable to that of today.
The Earth in the Late Triassic.
After the Triassic Sanromá and Palle moved on to the Mississippian (Early Carboniferous), 340 million years ago. During this period the continent of Pangea was coming together, though the continents were still largely separate. The climate was warmer that today, with inland seas on many continents, but there was still glaciation at the South Pole. The period is noted for extensive forests that covered much of the land masses. The model that Sanromá and Palle constructed of the period has a photometric variability of 4.46 and an albedo of 0.329; again not greatly dissimilar to today.
The Earth in the Mississippian.
Finally Sanromá and Palle constructed a model of the Late Cambrian, 500 million years ago. During the Cambrian a global supercontinent, Pannotia, had started to break up, with three island continents, Laurentia, Baltica and Siberia (roughly analogous to North America, Europe and Asia) and a residual Supercontinent, Gondwana, mad up of the remaining continental plates. The Cambrian had a warm climate, but with some glaciation at the poles. Most importantly, the Cambrian was before the evolution of vascular plants. It is thought that algae, fungi and lichens colonized the land some time before vascular plants, though it is unclear how early. For the sake of the model Sanromá and Palle assumed the Earth's land masses to be lacking vegetation of any sort during the Late Cambrian; this may not be completely accurate, but it is clear that at some point the Earth's landmasses did lack vegetation, so this model has some use. This model produced an albedo of 0.351, not greatly different to that of later periods, but a photometric variability of 12.2% which is distinctive.
The Earth during the Late Cambrian.
It is predicted that within the next few years we will have the technology to detect Earth-sized planets orbiting other stars, and therefore potentially to detect other planets with the capability to support life. However an Earth-like planet supporting life is not necessarily a familiar place; the Earth had unicellular life (bacteria, algae etc.) in its oceans for billions of years before the emergence of multicellular forms such as animals and plants, and after these emerged it took time for them to colonize the land. Sanromá and Palle's models suggest that it would be possible for a telescope to tell the difference between an Earth-like planet with plant cover and one without, on the basis of its photometric variability.
See also The Kepler-18 planetary system, Just how big is Iota Draconis b? The End of the Cretaceous and Exoplantes on Sciency Thoughts YouTube.